Impact of Physical Inactivity on Adipose Tissue Low-Grade Inflammation in First-Degree Relatives of Type 2 Diabetic Patients

The data are part of a larger study on the influence of physical inactivity in healthy and prediabetic subjects as initiated and funded within the framework program of the European Union EXGENESIS (Health Benefits of Exercise: Identification of Genes and Signalling Pathways Involved in Effects of Exercise on Insulin Resistance, Obesity and the Metabolic Syndrome) consortium.

The study participants were 33 young healthy men and included 13 FDRs of patients with type 2 diabetes and 20 CON subjects without any family history of diabetes. The groups were matched with respect to age, BMI, and cardiorespiratory fitness. The purpose and potential risk of the study were explained to subjects before they gave written informed consent. The study was approved by the ethical committee of Copenhagen and Frederiksberg, Denmark (protocol no. [01]-262546). Inclusion criteria were male, white, age 20–30 years, fasting plasma glucose <6.1 mmol/L, Vo_2max 35–55 mL/kg/min, and BMI 18.5–24.9 kg/m².

Fasting plasma glucose was determined in capillary blood using an ABL 625 (Radiometer, Copenhagen, Denmark). Body composition was determined by dual-energy X-ray absorptiometry full-body scanning (DPX-IQ 4.7e, Lunar Radiation Corporation, Madison, WI). Vo_2max was measured using a bicycle ergometer exercise protocol by means of an Oxycon Pro System (Jaeger, Höchberg, Germany).

All subjects participated in one experimental day approximately 3 weeks before 10 days of strict bed rest (BR). On day 10 of the BR period, subjects participated in a second and identical experimental day. Four days before the first experimental day and during the BR period, subjects were provided a standardized isocaloric diet (55 E% carbohydrates, 15 E% protein, 30 E% fat) and wore an Actiheart, a combined accelerometer and heart rate sensor (Cambridge Neurotechnology, Cambridge, U.K.). Subjects were instructed to continue their usual activities of daily living during the 4 days and in the 3 weeks before the BR period and to refrain from vigorous physical activity 24 h before the first experimental day. During the BR period, subjects remained in bed all day under surveillance.

The experimental day consisted of a 210-min baseline period, followed by a 180-min hyperinsulinemic (40 mU/min/m², Actrapid, Novo Nordisk, Copenhagen, Denmark) euglycemic clamp, as described previously (9). Microdialysis catheters were inserted and microdialysate was sampled during the last 60 min of baseline and the last 60 min of the clamp period. Subcutaneous adipose tissue blood flow was measured continuously, as previously described (10). An arterial catheter was inserted in the brachial artery for blood sampling. A venous catheter was inserted in the medial antecubital vein for infusion of insulin and glucose.

Large-pore microdialysis probes were manufactured in the laboratory from semipermeable hollow fibers (Plasmaflo OP-05, Asahi Medical Co., Tokyo, Japan) with a molecular mass cutoff >950 kDa (11). An 18-gauge cannula was used to insert four microdialysis probes into the subcutaneous abdominal adipose tissue (SCAAT) and four probes into subcutaneous femoral adipose tissue (SCFAT), as described previously (11). The probes were perfused with Ringer acetate containing glucose (2 mmol/L) using a high-precision CMA100 syringe pump (CMA Microdialysis AB, Solna, Sweden) at a rate of 1 μL/min.

Blood samples for analysis of adipokines and CRP were distributed in iced tubes containing 1.5 mg EDTA, and for analysis of insulin, the tubes contained 500 Kallikrein inhibitor units Trasylol and 15 mg of EDTA per mL of blood.

Plasma and microdialysate adiponectin were measured by a human radioimmunoassay kit (Linco Diagnostics Inc., St. Charles, MO). The intra- and interassay coefficients of variation were 9.3 and 1.8%. Plasma and microdialysate IL-10, IL-6, tumor necrosis factor-α (TNF-α), monocyte chemoattractant protein-1 (MCP-1), and leptin were analyzed by solid-phase protein immunochemical assays (Luminex 100IS, Ramcon A/S, Birkeroed, Denmark). Microdialysate IL-10 was analyzed using a LINCOplex kit (Human Cardiovascular Disease, Panel 3, HCVD3–67CK, Linco Research, St. Charles, MO). The intra- and interassay coefficients of variation were 8.3 and 11.2%. Microdialysate IL-6 and TNF-α as well as microdialysate and plasma leptin and MCP-1 were analyzed using a MILLIPLEX MAP kit (Human Serum Adipokine, Panel B, HADK2–61K-B, Millipore Corp., St. Charles, MO). The intra- and interassay coefficients of variation were 7.8 and 18.0% for IL-6, 7.9 and 15.0% for leptin, 7.8 and 16.0% for TNF-α, and 7.9 and 18.0% for MCP-1. Plasma TNF-α, IL-6, and IL-10 were analyzed using a MILLIPLEX MAP kit (High Sensitive Human Cytokine Kit, HSCYTO-60SPMX13, Millipore Corp.). The intra- and interassay coefficients of variation were 3.5 and 3.8% for TNF-α, 3.5 and 4.5% for IL-6, and 3.3 and 11.8% for IL-10. Plasma insulin was analyzed by ELISA technique (DAKO Diagnostics, Cambridgeshire, U.K.). Plasma CRP was measured with a CRP (Latex) HS Kit (Roche, Mannheim, Germany).

SCAAT biopsy specimens were collected in the basal state and at the end of a 180-min hyperinsulinemic (80 mU/min/m²) euglycemic clamp before BR and on day 9 of the BR intervention using the Bergstrom biopsy needle technique. Specimens were frozen in liquid nitrogen and stored at −80°C.

Total RNA was isolated using Trizol reagent (Gibco BRL, Life Technologies, Roskilde, Denmark), and cDNA was made with random hexamer primers using the GeneAmp PCR kit (Applied Biosystems, Carlsbad, CA). GAPDH was chosen as housekeeping gene. Quantification was performed with a SYBR Green real-time PCR assay, as previously described (12).

Data were analyzed using a linear mixed model where results are adjusted for variation within individuals. The model estimates differences between mean values. SAS 9.1 software (SAS Institute, Inc., Cary, NC) was used for statistical analysis.

Figures 1 and 2 show absolute means ± SE, and the differences and 95% CI described in the 15results are from the linear mixed model. If data were not normally distributed or homogeneous, as assessed by residual plots of each dependent variable, log-transformed data were used in the linear mixed model and significant effects described as relative (%) differences.

We evaluated the effect of groups (CON and FDR), BR (before vs. after), and time (baseline or insulin stimulated), and tested for differential effects of BR between the FDR and the CON groups. We also evaluated regional differences in microdialysate fluid proteins (SCAAT vs. SCFAT) at baseline before and after BR. Threshold for significance was P ≤ 0.05. The Bonferroni corrected P value that is equivalent to an uncorrected P = 0.05 in the current study is P = 0.0028, although this correction is likely overly conservative.

The FDR group was characterized by increased body fat percentage compared with the CON group (25.0 ± 2.3% vs. 17.4 ± 1.7%, P < 0.05) and increased abdominal adiposity (trunk/total fat mass: 0.58 ± 0.001 vs. 0.48 ± 0.1, P < 0.05) but similar BMI (24.9 ± 0.9 vs. 24.1 ± 0.5 kg/m², P > 0.05). In response to BR, no changes in anthropometrics were observed. Total, resting, and activity-related energy expenditure, as well as the physical activity score on Actiheart recordings, did not differ between groups during daily living and decreased in both groups during BR, as described previously (10). Vo_2max did not differ between FDR and CON groups before BR (39.1 ± 1.4 vs. 43.5 ± 1.3 mL O₂/kg/min, P > 0.05) and did not change in response to BR.

Before BR, whole-body insulin sensitivity (M value) was lower in the FDR compared with the CON group (4.3 ± 0.5 vs. 6.8 ± 0.5 mg/min/kg, P < 0.05), and BR decreased the M value in both FDR (P = 0.007) and CON subjects (P = 0.0001), as described previously (10).

SCAAT blood flow did not differ between groups and did not change in response to BR or insulin stimulation. SCFAT blood flow was lower before BR in FDR subjects compared with CON subjects during insulin stimulation (data not shown). SCFAT blood flow did not change in response to BR or insulin stimulation.

Levels of basal plasma leptin (Fig. 1D), MCP-1 (Fig. 1E), CRP, and insulin were higher in FDR than in CON subjects (difference between groups: leptin: 5,369 pg/mL [95% CI 1,031–9,708], P = 0.02; MCP-1: 29.3 [9.0–49.6], P = 0.006; CRP: 0.68 mg/dL [0.26–1.10], P = 0.003; and insulin: 15.6 pmol/L [2.1–29.0], P = 0.02).

Plasma adiponectin decreased in FDR subjects in response to BR (decrease: 1,060 pg/mL [95% CI 81–2,038], P = 0.04 [basal]; 692 pg/mL [20–1,365], P = 0.04 [insulin-stimulated]) and increased in CON (increase: 825 pg/mL [32–1,619], P = 0.04 [insulin-stimulated]). Basal plasma TNF-α and IL-10 increased in response to BR only in FDR subjects (increase: 2.4 pg/mL [0.52–4.32], P = 0.02; 54% [12–113], P = 0.01, respectively).

SCFAT microdialysate IL-6 was lower in the FDR compared with the CON group (difference between groups: 399.8 pg/mL [95% CI 32–768], P = 0.03 [basal, before BR]), whereas SCAAT microdialysate IL-10 was lower in the FDR compared with the CON group (difference between groups: 17.6 pg/mL [0.52–34.6], P = 0.04 [insulin-stimulated, before BR]; Table 1).

In response to BR, basal SCAAT microdialysate leptin decreased in CON subjects (decrease: 64% [95% CI 20–84], P = 0.01). In CON subjects, but not in FDR subjects, SCAAT microdialysate IL-10 increased during insulin stimulation (increase: 14.9 pg/mL [5.5–24.3], P = 0.004 [before BR]; 18.9 pg/mL [9.7–28.2], P = 0.0004 [after BR]).

In CON subjects, microdialysate leptin was higher in SCFAT than in SCAAT (difference: 314 pg/mL [95% CI 84–545], P = 0.01 [before BR]; 214 pg/mL [28–399], P = 0.03 [after BR]; Table 1). In FDR subjects, basal microdialysate IL-6 was higher in SCAAT than in SCFAT before BR (difference: 636 pg/mL [257–1,015], P = 0.003).

Adiponectin (Fig. 2A) and leptin (Fig. 2D) expressions were lower in the FDR compared with the CON group (difference between groups: adiponectin: 93% [95% CI 31–99], P = 0.001 [basal]; 86% [53–96], P = 0.002 [insulin-stimulated]; leptin: 86% [47–96], P = 0.006 [basal]; and 84% [33–96], P = 0.01 [insulin-stimulated]). IL-6 (Fig. 2B) and MCP-1 (Fig. 2E) expressions were higher in the FDR compared with the CON group (difference between groups: IL-6: 75% [42–89], P = 0.002 [basal]; 56% [7–79], P = 0.03 [insulin-stimulated]; and MCP-1: 0.0009 [0.0002–0.0016], P = 0.02 [basal]).

In CON subjects, basal adiponectin expression increased in response to BR (increase: 45% [95% CI 2–69], P = 0.04). In the FDR group, basal IL-10 expression decreased in response to BR (decrease: 77% [24–93], P = 0.02) and adiponectin and IL-10 expressions increased upon insulin stimulation after BR (increase: adiponectin: 97% [80–99.5], P = 0.0009; IL-10: 75%, [13–93], P = 0.03). In CON subjects, leptin and MCP-1 expression increased in response to insulin stimulation before BR (increase: leptin: 49% [14–70], P = 0.02; MCP-1: 0.0025 [0.0002–0.0048], P = 0.04).

The key finding is that nonobese, insulin-resistant individuals with a family predisposition for type 2 diabetes exhibit low-grade inflammation and, importantly, as little as 10 days of physical inactivity negatively affected the condition. The FDR and CON groups both exhibited decreased whole-body insulin sensitivity in response to BR.

The anti-inflammatory, antidiabetic, and antiatherogenic properties of adiponectin are mediated through a reciprocal association between adiponectin and CRP (13). A low adiponectin expression, as in our FDR subjects, may play a pathogenetic role in the development of insulin resistance and is probably the result of an increased level of CRP and proinflammatory cytokines. Indeed, we found higher plasma CRP and IL-6 gene expression in adipose tissue in FDR subjects than in CON subjects. TNF-α is a potent stimulator of IL-6 expression and secretion (1) and a potent inhibitor of adiponectin (14). The higher plasma TNF-α, as demonstrated in our FDR subjects, may partly explain the higher gene expression of SCAAT IL-6 and the lower gene expression of adiponectin. Because TNF-α and IL-6 are both known to reduce the expression of insulin receptor substrate-1, GLUT-4, and peroxisome proliferator–activated receptor-γ in adipocytes as well as decrease insulin-stimulated glucose transport (1,15), these factors may hold important functions for the insulin resistance found in FDRs.

Failure of adipocytes to adapt to energy storage is associated with reduced production of adiponectin and increased secretion of cytokines such as TNF-α (16,17), thus linking metabolic homeostasis and inflammation and suggesting that adipogenesis is pivotal for modulation of inflammation, which was also previously demonstrated by Menghini et al. (18). Increased TNF-α and decreased adiponectin, as in our FDR subjects, suggests that adipogenesis in these individuals is affected and plays a central role for the inflammatory condition.

Higher plasma concentrations of leptin have been found in insulin-resistant FDR subjects, and a positive correlation between serum leptin and body fat percentage has been found in both FDR and CON subjects (19); hence, the higher plasma leptin in our FDR subjects could simply be explained by their ∼50% higher fat mass. Confirming previous findings (11), we found higher microdialysate leptin in SCFAT compared with SCAAT in CON subjects.

MCP-1 recruits monocytes, inducing macrophage infiltration in adipose tissue, and even hypophysiologic levels of MCP-1 induced robust insulin resistance in human skeletal muscle cells (20). Interestingly, we found higher circulating MCP-1 and higher SCAAT expression of MCP-1 in FDR subjects than in CON subjects, suggesting an involvement of this adipokine in insulin resistance in FDRs.

The effect of physical inactivity on plasma TNF-α was pronounced in FDR subjects. Adipose tissue macrophages are a major source of TNF-α, which could indicate that macrophages infiltrate adipose tissue and secrete TNF-α in response to physical inactivity. Adipose tissue from obese individuals contains a higher number of macrophages than from lean subjects, and we previously demonstrated that a 15-week lifestyle intervention reduced the expression of macrophages and inflammatory proteins in adipose tissue of severely obese subjects (12).

Participants of the current study did not gain fat mass during the BR period; hence, the increase in plasma TNF-α must be ascribed to physical inactivity per se. Plasma TNF-α is the soluble fraction generated from membrane TNF-α caused by TNF-α converting enzyme (TACE) activity. TACE activity in the current study was measured indirectly by its substrates soluble intracellular adhesion molecule 1 and soluble vascular cell adhesion molecule 1 (data not shown), and both were increased by BR only in FDR subjects, supporting previous findings of a linkage between TACE activity and TNF-α (21).

A BR-induced decrease in plasma adiponectin, as found in the FDR subjects, could impair adipocyte differentiation (22) and induce enlargement of adipocytes and development of obesity during prolonged periods of physical inactivity. However, SCAAT expression of adiponectin and plasma adiponectin were upregulated in CON individuals in response to BR, indicating a compensatory action to the BR-induced reduction in whole-body insulin sensitivity. Interestingly, plasma IL-10 was upregulated in FDR subjects in response to BR. IL-10 modulates intramuscular fatty acid–derived metabolites and inhibits proinflammatory cytokine secretion and action (23,24). Thus, elevation of plasma IL-10 could be a compensatory mechanism in FDR subjects in response to the attenuated low-grade inflammation after BR. SCAAT microdialysate leptin was downregulated in CON subjects in response to BR, which could be explained by a decrease in adipose tissue lipid turnover (25). To maintain homeostasis, these adaptations are rational, but whether the adaptations persist if the period of physical inactivity is prolonged is unknown.

In conclusion, these data demonstrate that nonobese, insulin-resistant individuals with a family disposition to type 2 diabetes exhibit low-grade inflammation indicating dysregulation of adipose tissue. An equally important finding was that as little as 10 days of physical inactivity severely deteriorated the condition of low-grade inflammation in FDR subjects. The findings emphasize a particular importance of avoiding even short periods of physical inactivity in people at increased risk of type 2 diabetes.

Financial support was received from a European Union grant (6th Framework LSHM-CT-2004-005272, EXGENESIS) and the Nordea Foundation. L.H. was granted a Ph.D. scholarship from the Ministry of Science, Technology and Innovation, Copenhagen, Denmark.

No potential conflicts of interest relevant to this article were reported.

L.H. wrote the manuscript. M.P.S., A.C.A., and N.B.N. researched data and reviewed and edited the manuscript. F.D. and A.V. contributed to discussion and reviewed and edited the manuscript. J.M.B. reviewed and edited the manuscript. B.S. wrote the manuscript.

The authors thank Regitze Kraunsøe, Jeppe Bach, Thomas Beck, Lenette Pedersen, and Pia Hornbek, as well as the laboratory at the Steno Diabetes Centre for technical assistance. The authors thank the metabolic kitchen at Steno Diabetes Hospital for providing and managing the diet, and the Steno Diabetes Centre for providing the location for the BR.